The prognosis of neuroblastoma, one of the most common childhood cancers, has not improved over the last two decades. It is almost universally fatal in older children with disseminated disease. In contrast, neuroblastoma of infants younger than one year may spontaneously remit, or it is frequently controlled and cured with current therapy. The long-term goal of this proposal is to critically define these factors and to determine their role in the clinical characteristics and therapeutic response of childhood neuroblastoma. Specifically we will determine if lymphocyte or serum mediated cytotoxicity against neuroblasts is due to normal differentiation antigens or tumor specific antigens, compare the cytotoxicity of bone marrow, lymph node and blood lymphocytes against autologous or allogeneic neuroblasts, and establish if alpha-fetoprotein, detected on established neuroblastoma cell lines, is also expressed on viable neuroblasts from patients and investigate the effect of neuroblast-associated alpha-fetoprotein on host immune response. Systematic studies of these parameters should establish the biological differences between neuroblastoma of younger and older children and provide the knowledge required for the development of effective therapy.